This master’s thesis examines the suitability of AI systems for translating cultural realia in tourism brochures. The aim of the study is to compare the quality of translations generated by ChatGPT, Gemini, and DeepL with those produced by master’s students in translation studies, and to analyze the extent to which AI systems can appropriately render culture-specific expressions. The theoretical section explains key concepts in translation studies, cultural realia, translation strategies, and the role of presuppositions and translators in the translation process. The empirical section is based on an analysis of fifteen cultural realia from German-language tourism brochures for the city of Graz (Austria), which were translated into Slovenian by the three AI systems as well as by the students. The translations were evaluated comparatively in terms of their appropriateness and the translation strategies employed. The results show that AI systems are already frequently able to provide adequate translation solutions, but their performance has limitations, particularly when it comes to highly culture-specific elements. Overall, the translations produced by the students prove to be more consistent and better tailored to the target audience. The study thus confirms that human translators, thanks to their linguistic, cultural, and subject-matter expertise, continue to play a central role in the translation of cultural realia, while AI systems can be viewed primarily as supportive tools in the translation process.
|
